Contrary to current recommendations for people with non-insulin dependent diabetes mellitus (NIDDM), several recent studies have showed metabolic benefits from meal frequency greater than three/day. Since current nutritional therapy generally fails, this has great potential. We propose extending the time-hereto-fore studied, to two weeks, comparing three meals a day to six, with a narrowly focused group, overweight people with NIDDM but still secreting insulin. Design is random crossover. We hope to find metabolic benefits with six meals a day.